1. Field of the Invention
The present invention relates generally to a magnetic anisotropy multilayer used in a magnetic random access memory, and more particularly, to a magnetic anisotropy multilayer including a plurality of CoFeSiB/Pt layers.
2. Description of the Related Art
A magnetic random access memory (hereinafter, referred to as a magnetic memory) is being used in military applications such as a missile, a spacecraft, and the like. The magnetic memory has advantages of volatile devices such as a dynamic random access memory (DRAM) and a static random access memory (SRAM), i.e., both high integration degree of a DRAM and high-speed performance of an SRAM. In addition, the magnetic memory has lower power consumption than a nonvolatile type flash memory, and it has great number of repetition times of record reproduction. Therefore, it is considered as a substitution for the existing memory that has been used in a mobile phone, a computer and a network. Also, attempts are being made to apply the magnetic memory to a radio frequency identification (RFID) tag requiring low price and volatility, and further there is a great likelihood that it can be applied to a robot for factory automation.
The magnetic memory is a magnetic memory device having magnetic tunnel junctions (MTJs) based on tunneling magnetoresistance (TMR). The magnetic memory can input data by using spin directions caused by self-revolution of electrons in the device. In detail, the resistance of the magnetic memory is changed according as the spin directions of adjacent magnetic layers become parallel or anti-parallel, and the spin direction can be controlled to be parallel or anti-parallel by applying a magnetic field from the exterior. By using this property, it is possible to input data in the magnetic memory.
In general, the MTJ is configured in the shape of a sandwich where an insulating layer (generally, Al2O3 or MgO layer) as a tunneling barrier is interposed between two ferromagnetic layers. A current flows perpendicular to each layer. One of the two ferromagnetic layers is a pinned layer acting as a reference layer and the other one is a free layer for magnetic recording or sensing. In case that the spin directions of the two ferromagnetic layers are parallel with each other when current flows, the resistance becomes small so that the tunneling probability of current becomes great. On the contrary, when the spin directions of the two ferromagnetic layers are anti-parallel, the resistance becomes large, which results in reducing the tunneling probability of current. For ultra high integration of the magnetic memory, it should be necessary to form submicron memory cells. If a unit MTJ shrinks in size and an aspect ratio of the cell is also reduced for achieving the high integration degree of the magnetic memory, multi-domains or a vortex is formed inside a magnetic substance of the MTJ due to strong diamagnetic field. This leads to an unstable cell-switching phenomenon, which decreases a writing margin.
When fabricating the cell with high aspect ratio, such a multi-domain structure is not formed in virtue of shape magnetic anisotropy but it is difficult to achieve high integration. Moreover, this requires high switching magnetic field so that it is impossible to highly integrate the device after all.
To overcome such a problem, a perpendicular magnetic anisotropy MTJ has been developed (Naoki Nishimura et al., J. Appl. Phys., vol. 91, p. 5246. 2002). In Nishimura et al., the MTJ was fabricated using rare-earth and transition-metal alloys such as TbFeCo and GdFeCo, which has been well known as perpendicular magnetic anisotropy material, as a free layer and a pinned layer, respectively. The magnetoresistance ratio of this MTJ was 55%. In addition, it was confirmed that there was no susceptibility distortion at the perpendicular magnetic anisotropy MTJ through a magnetic force microscope (MFM). However, Tb and Gd used in this experiment are disadvantageous in terms of low corrosion resistance and difficulty in property control, and thus it is not easy to put these elements into the practical use. Therefore, for practical use of the perpendicular magnetic anisotropy MTJ (in short, pMTJ), it is necessary to develop new perpendicular magnetic anisotropy material.
The perpendicular magnetic anisotropy layers that have been researched were developed in order that they may be substituted for longitudinal magnetic storage media which will encounter the limitation of high density. The material exhibiting the perpendicular magnetic anisotropy is CoCr-based alloy layer, Co/Pt, or Co/Pd multilayer, wherein this material should meet specific physical properties such as high perpendicular magnetic anisotropy, high coercivity and high remanent magnetization.
However, as the magnetic memory needs rapid switching and low power consumption, low coercivity and high magnetic anisotropy for the increase of a writing margin are required. In addition, it is required to maintain a remanent magnetization to be similar to a saturation magnetization and simultaneously to be low for improving the sensitivity of the magnetic memory switching operation.
In order to employ a perpendicular magnetic tunnel junction (pMTJ) that has been actively researched to meet the demand for high integration of the magnetic memory, the perpendicular magnetic anisotropy layer exhibiting low coercivity, low saturation magnetization and high magnetic anisotropy should be used. That is, the perpendicular magnetic anisotropy MTJ formed of the above-listed perpendicular magnetic anisotropy material, in which the magnetization direction is perpendicular to the layer surface, has low saturation magnetization and no magnetic distortion at an edge of the layer, which will make it possible to realize the high integration of the magnetic memory.
Accordingly, it is required a perpendicular magnetic anisotropy layer having a low coercivity and a low saturation magnetization similar to a remanent magnetization, and capable of minimizing power consumption as well.